Targeted Modification of furan-2-carboxaldehydes into Michael Acceptor Analogs Yielded Long-Acting Hemoglobin Modulators with Dual Antisickling Activities

Abstract

Sickle cell disease (SCD) is the most common genetic disorder, affecting millions of people worldwide. Aromatic aldehydes, which increase the oxygen affinity of human hemoglobin to prevent polymerization of sickle hemoglobin and inhibit red blood cell (RBC) sickling, have been the subject of keen interest for the development of effective treatment against SCD. However, the aldehyde functional group metabolic instability has severly hampered their development, except for Voxelotor, which was approved in 2019 for SCD treatment. To improve the metabolic stability of aromatic aldehydes, we designed and synthesized novel molecules by incorporating Michael acceptor reactive centers into the previously clinically studied aromatic aldehyde, 5-hydroxymethylfurfural (5-HMF). Eight such derivatives, referred to as MMA compounds were synthesized and studied for their functional and biological activities. Unlike 5-HMF, which forms Schiff-base interaction with αVal1 nitrogen of hemoglobin, the MMA compounds covalently interacted with βCys93, as evidenced by reverse-phase HPLC and disulfide exchange reaction, explaining their RBC sickling inhibitory activities, which at 2mM and 5 mM, ranges from 0–21% and 9–64%, respectively. Additionally, the MMA compounds showed a second mechanism of sickling inhibition (12–41% and 13–62% at 2mM and 5mM, respectively) by directly destabilizing the sickle hemoglobin polymer. In vitro studies demonstrated sustained pharmacologic activities of the compounds compared to 5-HMF. These findings hold promise for advancing SCD therapeutics.

Introduction

Sickle cell disease (SCD) poses a significant public health burden, with health costs exceeding $1 billion annually in the United States alone.(Hassell, 2010; Kauf et al., 2009; Piel et al., 2013) The disease occurs as a result of a single nucleotide mutation, leading to the amino acid change from glutamic acid to valine residue at the sixth position of the beta-globin chain (βGlu6 in normal hemoglobin, HbA to βVal6 in sickle Hb, HbS). When HbS is deoxygenated, it polymerizes into long and rigid fibers that lead to sickling of red blood cells (RBCs.).(Akinsheye & Klings, 2010; Aliyu et al., 2008; Belcher et al., 2003; De Franceschi, 2009; Piel et al., 2017; Safo et al., 2021) The hypoxia-induced sickling causes several downstream pathophysiological events that include red blood cell hemolysis, impaired microvascular blood flow, inflammation, oxidative stress, decreased vascular NO bioavailability, vaso-occlusion, painful crises, morbidity, and premature mortality. It was recognized several decades ago that therapies that prevent the hypoxia-induced polymerization are likely to have a beneficial impact on the subsequent pathophysiology of SCD.(Abraham et al., 1991; Arya et al., 1996; Beddell et al., 1984; Keidan et al., 1986; Merrett et al., 1986; Zaugg et al., 1977) Consequently, several compounds, most notably aromatic aldehydes that bind to Hb and destabilize the low-O₂-affinity T-state and/or stabilize the high-O₂-affinity R-state to increase the concentration of the non-polymer forming oxygenated HbS have been investigated for the treatment of SCD.(Abdulmalik et al., 2005, 2011, 2020; Ahmed et al., 2020; Deshpande et al., 2018; Dufu & Oksenberg, 2018; Metcalf et al., 2017; Oder et al., 2016; Pagare et al., 2018, 2022; Safo et al., 2004, 2011, 2021; Safo & Kato, 2014) Except Voxelotor (GBT440; Figure 1), which was approved in 2019 by the FDA for the treatment of SCD,(Metcalf et al., 2017; Oksenberg et al., 2016; Vichinsky et al., 2019) other aromatic aldehydes, despite promising results in early clinical trials,(Abdulmalik et al., 2005; Oder et al., 2016; Safo et al., 2004; Safo & Kato, 2014) ultimately failed because of poor pharmacokinetic properties as a result of metabolic instability of the aldehyde moiety.(Abdulmalik et al., 2005; Abraham et al., 1991; Oder et al., 2016; Safo et al., 2004; Safo & Kato, 2014; Vasiliou et al., 2000; Yoshida et al., 1998) The ability of aromatic aldehydes to increase Hb affinity for oxygen and the consequent inhibition of RBC sickling depends on their property to engage in Schiff-base interaction with the N-termini αVal1 amines of the Hb α-subunits.(Abdulmalik et al., 2005, 2011; Ahmed et al., 2020; Deshpande et al., 2018; Dufu & Oksenberg, 2018; Metcalf et al., 2017; Oder et al., 2016; Oksenberg et al., 2016; Pagare et al., 2018; Safo et al., 2004, 2011; Safo & Kato, 2014; Vichinsky et al., 2019) However, as noted earlier, aromatic aldehydes have the propensity to undergo rapid in vivo oxidative metabolism, which limits the bioavailability of many promising aromatic aldehyde drug candidates.(Abdulmalik et al., 2005; Abraham et al., 1991; Godfrey et al., 1999; Obied &, T., Venitz, J, 2009; Safo et al., 2004; Vasiliou et al., 2000; Yoshida et al., 1998) Voxelotor, and other recently discovered aromatic aldehydes solve this metabolic instability by having ortho hydroxyl group on the benzene ring (relative to the aldehyde moiety) that provides protection against unwanted metabolic breakdown by enzymes.(Pagare et al., 2022)

Figure 1.

Examples of previously studied antisickling compounds

Several non-aromatic aldehyde compounds have also been studied for their RBC sickling inhibition potentials.(Kennedy et al., 1984; Nakagawa et al., 2014, 2018; Omar et al., 2015, 2016, 2019, 2020; Pagare et al., 2022; Park et al., 2003) These compounds, unlike aromatic aldehydes effect their RBC sickling inhibition activities by covalently binding to the surface-located βCys93 of Hb to destabilize the T-state and shift the allosteric equilibrium to the high-O₂-affinity R-state, resulting in increased Hb oxygen affinity. This class of compounds are also expected to directly destabilize the HbS polymer by binding to the Hb surface, and stereospecifically prevent HbS molecules from coming close together. An example of βCys93 binder is Ethacrynic acid (Figure 1), with a terminal enone reactive moiety.(Kennedy et al., 1984) However, potential toxicity and the diuretic activity of Ethacrynic acid prevented its development for the treatment of SCD. Other examples are thiol and the azolylacryloyl derivatives KAUS-15 and KAUS-38 (Figure 1).(Nakagawa et al., 2014, 2018; Omar et al., 2015, 2019, 2020)

5-Hydroxymethylfurfural (5-HMF; Figure 1) was one of the most promising antisickling aromatic aldehydes that failed in phase I/II clinical studies for the treatment of SCD due in part to metabolic instability of the aldehyde.(Abdulmalik et al., 2005; Oder et al., 2016; Pagare et al., 2022; Safo et al., 2004, 2021; Safo & Kato, 2014) Several aromatic aldehyde analogs of 5-HMF, e.g. 5-PMFC that showed improved sickling inhibition, also had similar metabolic liability.(Xu et al., 2017) In this study, we have replaced the aldehyde moiety of 5-HMF and 5-PMFC with Michael acceptor (MA) reactive moiety. We hypothesized that this replacement will preserve their sickling inhibition potency, and improve their in vitro pharmacokinetic properties. Eight compounds were designed, synthetized, and tested for their in vitro pharmacodynamic (PD) effects, specifically, Hb modification, increased Hb oxygen affinity, and antisickling activity. Their mode(s) of interaction with Hb and, consequently, their mechanism(s) of action were also studied using reverse-phase HPLC and disulfide exchange reaction between DTNB and the thiol of βCys93 of Hb.

Materials and methods

Study approvals

Study approvals that involve blood obtained from volunteers at both Virginia Commonwealth University (VCU) and the Children’s Hospital of Philadelphia (CHOP) follow similar published procedures by our group(Omar et al., 2020). At VCU, normal whole blood (AA) was collected from adult donors (>18 years) after informed consent, in accordance with regulations of the IRB for Protection of Human Subjects (IRB #HM1) by the Institutional Review Board at VCU. At CHOP, leftover blood samples from individuals with homozygous sickle cell (SS) who had not been recently transfused, were obtained and utilized based on an approved IRB protocol (IRB# 11-008151) by the Institutional Review Board, with informed consent. All experimental protocols and methods were performed in accordance with institutional (VCU and CHOP) regulations.

General procedure

Characterization of the synthetic organic compounds were characterized following previously published procedures(Omar et al., 2020). Briefly, melting point was performed using OptiMelt Automated Melting Point System Digital Image Processing Technology SRS, Stanford Research Systems (Sunnyvale, CA, USA); NMR spectroscopy was recorded on Bruker AVANCE III 400 (Bruker, Fällanden, Switzerland); LC-MS spectroscopy was performed on Agilent Technologies 1260 Infinity LC/MSD system with DAD\ELSD Alltech 3300 and Agilent; LC\MSD G6120B mass-spectrometer (Santa Clara, CA, USA); High-resolution mass spectroscopy (HRMS) was performed by separation and mass spectrometric detection techniques and were performed with an Infinity 1260 UHPLC system (Agilent Technologies, Waldbronn, Germany) coupled to an 6224 Accurate Mass TOF LC/MS system (Agilent Technologies, Singapore).

Chromatographic separation also followed published procedures(Omar et al., 2020), using Agilent Zorbax C18 column (100mm× 2.1mm, 1.9 μm particle size), and mobile phases A and B consisting of 0.1 % formic acid in water and 0.1 % formic in acetonitrile, respectively. Solvents and chemicals were HPLC grade. Water was purified by Millipore Water Purification System. All the chemicals, reagents and solvents were purified and/or dried in accordance to well-known literature methods; vendors’ names: UORSY and Enamine (Kiev, Ukraine). All reactions were performed under inert atmosphere unless stated otherwise.

Preparation of MMA-401, MMA-402 and MMA-403 (Scheme 1):

Scheme 1:

Synthetic scheme for MMA-401, MMA-402, and MMA-403.

(i) LiCl, DBU, MeCN, rt, 14 h. (ii) LiOH, THF/H₂O, rt, 7h. (iii) NH₂OH‧HCl, EtOH, H₂O, reflux, 20 min. (iv) TFAA, Et₃N, THF, 0 °C to rt, 12h.

Synthesis of methyl (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylate (3)

To a suspension of 2a, LiCl (0.2 g, 4 mmol) in dry MeCN (20 mL) was added 5-HMF (0.84 g, 4 mmol) and DBU (0.6 mL, 4 mmol). The mixture was stirred for 10 min at room temperature before diethyl (2-methoxy-2-oxoethyl)phosphonate (2a) (0.84 g, 4 mmol, in dry CHCl₃) was added. The reaction was monitored via TLC and quenched with saturated NH₄Cl after complete conversion. The organic solvent was removed, and the aq layer extracted with EtOAc (3×25 mL). The combined organic layers were washed with water (2×15 mL) and brine (10mL), dried over Na₂SO_4, evaporated in vacuo and purified by column chromatography (silica gel eluting with DCM/hexane, 1:1) to give methyl (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylate (3) (0.64 g, 88% yield); mp 66 °C. ¹H NMR (500 MHz, CDCl₃) δ 7.39 (d, J = 15.8 Hz, 1H), 6.56 (d, J = 3.5 Hz, 1H), 6.37 (d, J = 3.4 Hz, 1H), 6.30 (d, J = 15.8 Hz, 1H), 4.65 (s, 2H), 3.78 (s, 3H), 2.13 (s, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 167.54, 156.55, 150.73, 131.08, 115.74, 115.40, 110.20, 57.60, 51.70. HRMS (ESI), t_R = 4.467 min, m/z 183.0649 [M + H]⁺, formula C₉H₁₀O₄.

Synthesis of (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylic acid (MMA-401)

The ester 3 (0.64 g) and LiOH (0.43g, 3 eq.) were added to a mixture of THF and H₂O (7:3, 10 ml) and stirred for 7 h at rt, followed by addition of MTBE (10 mL). The aq layer was separated and acidified with aq 4 M HCl. Subsequently, the aq layer was extracted with EtOAc (3×10 mL). The combined organic layers were washed with brine (2×10mL), dried (with anhydrous Na₂SO₄), filtered, and evaporated to dryness under reduced pressure to afford (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylic acid (MMA-401) (0.35 g, 60% yield); mp 138 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 12.33 (s, 1H), 7.35 (d, J = 15.7 Hz, 1H), 6.85 (d, J = 3.3 Hz, 1H), 6.43 (d, J = 3.3 Hz, 1H), 6.10 (d, J = 15.7 Hz, 1H), 5.36 (t, J = 5.9 Hz, 1H), 4.42 (d, J = 5.4 Hz, 2H); ¹³C NMR (126 MHz, Chloroform-d) δ 169.08, 157.41, 150.48, 131.26, 115.90, 115.54, 109.77, 57.15, 40.62 – 39.26 (m); HRMS (ESI), t_R = 0.790 min, m/z 167.0350 [M - H]^-, formula C₈H₈O₄.

Synthesis of 5-(hydroxymethyl)furan-2-carbonitrile (MMA-402)

A mixture of 5-HMF (0.5 g, 3.9 mmol) in 50 mL EtOH and hydroxylamine hydrochloride (0.505 g, 7.1 mmol) in H₂O (5 mL) was heated to reflux for 20 min. The reaction mixture was cooled and quenched with aq HCl (10%) to neutral pH. The product was extracted with CH₂Cl₂ (2×20mL), and the combined organic phase was dried over Na₂S0₄, and concentrated in vacuo to give (E)-5-(hydroxymethyl)furan-2-carbaldehyde oxime (4) (0.297g, 2.1mmol, 53.84% yield).

To a solution of (E)-5-(hydroxymethyl)furan-2-carbaldehyde oxime (4) (0.564 g, 4 mmol) and 20 mL THF was added 4 mL of Et₃N, and the mixture cooled to 0 °C. To this mixture was added 2 mL TFAA dropwise, and stirred at rt for 12 h. The reaction was quenched with a saturated solution of NaHCO₃, stilled for 30 minutes, and extracted with CH₂Cl₂ (3×20mL). The organic layers were combined, washed consecutively with 5% HCl (20mL), water (2×15 mL), and brine (10 mL). The solution was concentrated in vacuo, and purified by column chromatography (silica gel eluting with DCM/hexane, 1:1) to give 5-(hydroxymethyl)furan-2-carbonitrile (MMA-402) as an oil (0.186 mg, 95+% purity, 37.8% yield). ¹H NMR (400 MHz, DMSO-d₆) δ 7.49 (d, J = 3.6 Hz, 1H), 6.56 (d, J = 3.6 Hz, 1H), 5.35 (d, J = 12.2 Hz, 1H), 4.47 (s, 2H).¹³C NMR (126 MHz, DMSO-d₆) δ 162.33, 124.89, 124.22, 112.40, 109.24, 56.06; HRMS (ESI), t_R = 2.075, m/z 124.0392 [M + H]⁺, formula C₆H₅NO₂.

Synthesis of (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylamide (MMA-403)

To a suspension of LiCl (0.2 g, 4 mmol) in dry MeCN (20mL) was added 5-HMF (0.84 g, 4 mmol) and 1,8-Diazabicyclo(5.4.0)undec-7-ene (DBU) (0.6 mL, 4 mmol). The mixture was stirred for 10 min at rt, followed by addition of diethyl (2-amino-2-oxoethyl)phosphonate (2b) (0.5g, 4 mmol, in dry CHCl₃). The reaction was monitored via TLC and quenched with saturated NH₄Cl after complete conversion (approx. 14–18 h). The organic solvent was removed, and the aq layer was extracted with EtOAc (3×25mL). The combined organic layers were washed with water (2×15 mL) and brine (10mL), dried over Na₂SO_4, evaporated in vacuo and purified by column chromatography (silica gel eluting with 1:1 DCM/hexane) to give (E)-3-(5-(hydroxymethyl)furan-2-yl)acrylamide (MMA-403) (0.36g, 2mmol, 46% yield); mp 131-156 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.53 (s, 1H), 7.17 (d, J = 15.7 Hz, 1H), 7.04 (s, 1H), 6.67 (d, J = 3.4 Hz, 1H), 6.36 (d, J = 15.2 Hz, 2H), 5.30 (t, J = 5.6 Hz, 1H), 4.41 (d, J = 5.8 Hz, 2H); ¹³C NMR (101 MHz, DMSO-d₆) δ 166.98, 157.76, 150.70, 127.08, 119.49, 115.10, 109.90, 56.27; HRMS (ESI), t_R =1.909 min, m/z 168.0654 [M + H]⁺, formula C₈H₉NO₃.

Preparation of MMA-407, MMA-408, MMA-409, MMA-410 and MMA-411 (Scheme-2):

Scheme 2:

Synthetic scheme for MMA-407, MMA-408, MMA-409, MMA-410, and MMA-411.

(i) DIAD, Ph₃P, phenol, THF, 0°C to rt, 3 h. (ii) RMgBr, THF, 0°C to rt, 10 h. (iii) 2-Iodoxybenzoic acid, EtOAc, reflux, 10 h. (iv) K₂CO₃, MeOH, rt, 1. (v) BuLi/hexanes, THF, CO₂, −80 °C. (vi) n-BuLi/hexanes, THF, −80 °C, then ethyl chloroformate, THF, −80 °C to rt. (vii) n-BuLi/hexanes, ZnCl₂, THF, AcCl, −80°C. (viii) ethyl chloroformate, Et₃N, THF, aq NH₃.

Synthesis of 4-(5-(phenoxymethyl)furan-2-yl)but-3-yn-2-one (MMA-408)

To a cooled (approx. 4 °C) solution of 5-HMF (3 g, 23.8 mmol), Ph₃P (6.89 g, 26.2 mol) and phenol (2.46 g, 2.62 mmol) in THF (20 mL), DIAD (5.29 g, 26.2 mmol) was added dropwise, and the mixture warmed to rt and stirred for 3 h. The solution was evaporated and purified by column chromatography (Hex/EtOAc, 4:1) to give 5-PMFC as an orange solid (16 g, 33.2% yield).

K₂CO₃ (1.05g, 7.2 mmol) was added to a solution of 5-PMFC (1g, 4.8 mmol) in dry methanol (10 mL) and cooled with ice to approx. 5°C. Following, a solution of dimethyl 1-diazo-2-oxopropylphosphonate (1.12 g, 5.94 mmol) in methanol (10 mL) was added dropwise, and the mixture stirred at rt for 11 h. After evaporating the solvent under reduced pressure, the product was reconstituted in CH₂Cl₂, washed with water and brine (10mL), dried over Na₂SO₄ and evaporated in vacuo to give crude 2-ethynyl-5-(phenoxymethyl)furan (7) (1 g, 94% yield), which was used without further purification.

To a solution of (7) (1 g, 5.05 mmol) in dry THF at −80°C was added dropwise 2.5 M solution of n-BuLi in hexane (2.1 mL, 5.55 mmol). The resulted solution was stirred for 30 min at −80°C. A solution of ZnCl₂ in THF was added to the mixture, followed by addition of acetyl chloride and stirred at the same condition. After 30 min, the mixture was quenched with NH₄Cl solution and extracted with EtOAc (20 mL×3). The organic layers were combined, washed with water (10 mL×2) and brine (10 mL), dried over Na₂SO₄ and concentrated in vacuum to give 1g of crude 4-(5-(phenoxymethyl)furan-2-yl)but-3-yn-2-one (MMA-408). After column chromatography (Hex:EtOAc, 7:1), MMA-408 was obtained as a yellow solid (0.48 g, 2 mmol, 26.5% yield); mp 90.3 – 91 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.34 – 7.26 (m, 3H), 7.03 (d, J = 8.0 Hz, 2H), 6.97 (t, J = 7.3 Hz, 1H), 6.80 (d, J = 3.7 Hz, 1H), 5.13 (s, 2H), 2.43 (s, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 155.77, 130.03, 123.75, 121.70, 115.21, 112.92, 61.82, 32.62. LC-MS (ESI), t_R = 1.425 min, m/z 241.2 [M + H]⁺; HRMS (ESI), t_R = 6.547 min, m/z 241.0862 [M + H]⁺, formula C₁₅H₁₂O₃.

Synthesis of 3-(5-(phenoxymethyl)furan-2-yl)propiolamide (MMA-407)

To a solution of (7) (1 g, 5.05 mmol) in dry THF that was cooled to −80 °C, 2.5 M solution of BuLi in hexane (2.1ml, 5.55mol) was added dropwise, and the mixture stirred for 30 min at −80 °C. The reaction flask was degassed and filled with dry CO₂, left to stir for 1h, and then slowly warmed to rt. After 30 min of further stirring, the mixture was quenched with NH₄Cl solution and extracted with EtOAc (3×20 mL). The organic layers were combined, washed with water (10mL×2) and brine (10mL), dried over Na₂SO₄ and evaporated in vacuo to give 1 g of crude 3-(5-(phenoxymethyl)furan-2-yl)propiolic acid (8), and used without further purification.

Ethyl chloroformate (0.497 g, 4.54 mmol) at −15°C to −10°C was added to a solution of (8) (1g, 4.1 mmol) and Et₃N (0.46g, 4.54 mmol) in THF (10mL), and the mixture stirred for 1h. Following, 3 mL of 25% aq ammonia was added and the reaction mixture stirred for 1h. The resulted mixture was diluted with EtOAc, washed 3 times with water (10mL), and the organic layer separated, dried over Na₂SO₄ and concentrated in vacuo. After column chromatography (Hex:EtOAc, 1:1) 3-(5-(phenoxymethyl)furan-2-yl)propiolamide (MMA-407) was obtained as a white solid (0.21g, 0.57mmol, 21.2% yield); mp 134.5-135 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 8.25 (s, 1H), 7.77 (s, 1H), 7.31 (t, J = 7.9 Hz, 2H), 7.07 (d, J = 3.5 Hz, 1H), 7.02 (d, J = 8.1 Hz, 2H), 6.97 (t, J = 7.3 Hz, 1H), 6.74 (d, J = 3.5 Hz, 1H), 5.10 (s, 2H). ¹³C NMR (400 MHz, DMSO-d₆) δ 158.25, 155.60, 154.34, 134.83, 129.18, 121.09, 120.22, 114.61, 110.9, 87.18, 74.86, 61.75. LC-MS (ESI), t_R =1.210 min, m/z 242.0 [M + H]⁺; HRMS (ESI), t_R = 0.790 min, m/z 242.0821 [M + H]⁺, formula C₁₄H₁₁NO₃.

Synthesis of methyl 3-(5-(phenoxymethyl)furan-2-yl)propiolate (MMA-411)

To a solution of (7) (1 g, 5.05 mmol) in dry THF at −80°C was added dropwise 2.5 M solution of BuLi in hexane (2.1 mL, 5.55mol). The resulted solution was stirred for 30 min at −80°C. Ethyl chloroformate (0.52g, 5.55mmol) in THF was then added to the reaction mixture and left to stir for 1 h, and slowly warmed to rt. After 30 min, the mixture was quenched with NH₄Cl solution and extracted with EtOAc (3×20mL). The organic layers were combined, washed with water (2×10mL) and brine (10mL), dried over Na₂SO₄ and evaporated in vacuo to give 1 g of methyl 3-(5-(phenoxymethyl)furan-2-yl)propiolate (MMA-411). After column chromatography (Hexanes/EtOAc, 4:1) MMA-411 was obtained as a yellow solid (0.148 g, 0.57 mmol, Yield = 11.5%); mp 60—63.4 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.38 – 7.23 (m, 3H), 7.04 (s, 2H), 7.00 – 6.93 (m, 1H), 6.79 (d, J = 3.5 Hz, 1H), 5.13 (s, 2H), 3.79 (s, 3H); ¹³C NMR (126 MHz, DMSO-d₆) δ 158.12, 155.65, 153.60, 133.61, 130.03, 123.61, 121.71, 115.22, 112.73, 86.37, 76.63, 61.79, 53.65. LC-MS (ESI), t_R = 4.09 min, m/z 257.0 [M + H]⁺; HRMS (ESI), t_R = 6.756 min, m/z 257.0816 [M + H]⁺, formula C₁₅H₁₂O₄.

Synthesis of 1-(5-(phenoxymethyl)furan-2-yl)prop-2-en-1-one (MMA-409)

To a solution of 5-PMFC (1 g, 4.95 mmol) in THF (10 mL) at 0 °C was added vinyl magnesium bromide (7.45 mL of 1N solution in THF, 7.45 mmol) and stirred at rt temperature for 10 h. The mixture was poured to EtOAc, washed with H₂O (2×15mL), dried over Na₂SO₄, and evaporated to obtain 1-(5-(phenoxymethyl)furan-2-yl)prop-2-en-1-ol (5a) (0.95 g, yield 83%), which was used in the next step without further purification.

2-Iodoxybenzoic acid (2.31 g, 8.26 mmol) was added to a solution of (5a) (0.95 g, 4.13 mmol) in dry EtOAc (20 mL), and then heated to reflux for 10 h. The mixture cooled, filtrated, and evaporated to give crude product 1-(5-(phenoxymethyl)furan-2-yl)prop-2-en-1-one (MMA-409). The pure product was obtained as a yellow solid (0.164 g, 0.71 mmol, 17.4% yield) after column chromatography (Hexanes/EtOAc, 4:1); mp 66.4 – 67.6 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.68 (d, J = 3.6 Hz, 1H), 7.31 (t, J = 8.0 Hz, 2H), 7.22 (dd, J = 17.0, 10.4 Hz, 1H), 7.05 (d, J = 8.1 Hz, 2H), 6.98 (t, J = 7.3 Hz, 1H), 6.85 (d, J = 3.6 Hz, 1H), 6.39 (dd, J = 17.1, 1.8 Hz, 1H), 5.94 (dd, J = 10.4, 1.8 Hz, 1H). ¹³C NMR (126 MHz, CDCl₃) δ 132.12, 130.04, 129.90, 121.72, 121.28, 115.21, 113.10, 62.05. LC-MS (ESI), t_R = 1.33 min, m/z 229.0 [M + H]⁺; HRMS (ESI), t_R = 6.132 min, m/z 229.0869 [M + H]⁺, formula C₁₄H₁₂O₃.

Synthesis of 1-(5-(phenoxymethyl)furan-2-yl)prop-2-yn-1-one (MMA-410)

To a solution of 5-PMFC (1 g, 4.95mmol) in THF (15 mL) at 0 °C was added ethynyl magnesium bromide (14 mL of 0.5N solution in THF, 7 mmol) and stirred at rt for 10h. The mixture was poured into EtOAc (20 mL), washed with H₂O (2×15 mL), dried over Na₂SO₄, and evaporated in vacuo to obtain 0.91 g of 1-(5-(phenoxymethyl)furan-2-yl)prop-2-yn-1-ol (5b) (yield = 81%) for use without further purification.

2-Iodoxybenzoic acid (2.23g, 7.98 mmol) was added to a solution of (17) (0.91 g, 3.99 mmol) in dry EtOAc (20 mL). The resulted suspension was refluxed for 11 h, cooled, filtrated and evaporated under vacuum to give a crude product. After column chromatography (Hex/EtOAc, 4:1), 1-(5-(phenoxymethyl)furan-2-yl)prop-2-yn-1-one (MMA-410) was obtained as a yellow solid (0.398 g, 1.76 mmol, 44.2%); mp 73.7 – 75.9 °C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.58 (d, J = 3.7 Hz, 1H), 7.36 – 7.25 (m, 2H), 7.04 (d, J = 8.0 Hz, 2H), 6.98 (t, J = 7.3 Hz, 1H), 6.88 (d, J = 3.7 Hz, 1H), 5.20 (s, 2H), 4.99 (s, 1H).¹³C NMR (126 MHz, DMSO-d₆) δ 163.56, 158.06, 157.82, 152.44, 130.05, 124.10, 121.79, 115.21, 113.57, 84.36, 80.20, 62.00. LC-MS (ESI), t_R = 1.362 min, m/z 227.2 [M + H]⁺; HRMS (ESI), t_R = 0.579 min, m/z 227.0712 [M + H]⁺, formula C₁₄H₁₀O₃.

In vitro Hemoglobin Modification, Oxygen Equilibrium and Antisickling Studies Using Human Homozygous Sickle cell (SS) Blood

The MMA series of compounds (MMA-401, MMA-402, MMA-403, MMA-407, MMA-408, MMA-409, MMA-410 and MMA-411) and the positive control 5-HMF were studied for their abilities to inhibit RBC sickling (RBC morphology study), increase Hb oxygen, and/or modify Hb (adduct formation) following previously established and published procedures by our group.(Abdulmalik et al., 2005, 2020; Pagare et al., 2018) Briefly, homozygous SS blood (hematocrit: 20%) suspensions were incubated under air in the absence or presence of 2 and/or 5 mM concentration of test compounds at 37° C for 1hr. Following, the suspensions were incubated under hypoxia (2.5% Oxygen gas/balance Nitrogen gas) at 37°C for 2 hr. Aliquot samples were fixed with 2% glutaraldehyde solution without exposure to air, and then subjected to microscopic morphological analysis. The residual samples were washed in phosphate-buffered saline, and hemolyzed in hypotonic lysis buffer for the Hb oxygen equilibrium and Hb adduct experiments. For the Hb oxygen equilibrium study, approximately 100 μl aliquot samples from individual lysates were added to 4 ml of 0.1M potassium phosphate buffer, pH 7.0, in a cuvette and subjected to hemoximetry analysis using Hemox^™ Analyzer (TCS Scientific Corp.) to assess P₅₀ shifts. Additional aliquots from the lysate were subjected to Hb adduct formation study by cation-exchange HPLC (Hitachi D-7000 Series, Hitachi Instruments, Inc., San Jose, CA) with a weak cation-exchange column (Poly CAT A: 35 mm x 4.6 mm, Poly LC, Inc., Columbia, MD). The binding of the MMA compounds to either α- or β-globin chains were also determined following published procedures(Omar et al., 2020). In this experiment, aliquot samples of the hemolysates from the above Hb adduct formation study were subjected to reversed-phase (RP) HPLC on a Hitachi D-7000 HSM Series, using a Jupiter 5 μm C-4 50 × 4.6 mm column, (Phenomenex^®, Torrence, CA) and a gradient from 20% to 60% acetonitrile in 0.3% trifluoroacetic acid in 15 minutes, with UV detection at 215 nm.

In another experiment to determine oxygen-independet antisickling effect of the compounds, which also follows our previously published procedure,(Abdulmalik et al., 2020) aliquotes from the above antisickling studies were fixed with 2% glutaraldehyde without exposure to air. Then the incubation chamber was opened and exposed to air for 15 minutes to ensure complete re-oxygenation and reversal of the sickled cells to normal round cells. Reversal was confirmed by microscopy. The incubation chamber was then closed, and the assay was repeated under 100% nitrogen gas for 30 minutes, at which point aliquots were obtained and fixed. Aliquot samples were then subjected to microscopic morphological analysis of bright field images as previously described.(Abdulmalik et al., 2020) Resulting sickled cells (percentages) were compared across samples, and between aliquots of the same samples that had been obtained either under 2.5% oxygen or 100% nitrogen.

Reactivity of MMA compounds toward βCys93 of Hb

The capability of the MMA compounds to react with Hb sulfhydryl groups were studied following similar procedures published by our group(Omar et al., 2015, 2019, 2020). Briefly, aqueous solution of Hb (50µM) was mixed with MMA 401, MMA 408, MMA 409 or MMA 410 (2 mM). Additionally, Hb was mixed with phosphate buffered saline (PBS) to serve as a negative control, and ethacrynic acid studied as a positive control. The mixtures were incubated for four hours at 25°C, and then centrifuged with washing to separate excess reagents from the Hb. The washed hemoglobin was stored overnight at 4°C. 5 µL of 100 mM DTNB was added to 25 µL of each Hb solution, and then diluted to 500 µL in 0.1 M phosphate buffer and incubated for an hour. Another 25 µL of each Hb solution was diluted to 500 µL in buffer to serve as a non-DTNB control. The tubes containing these solutions were centrifuged, and the collected yellow filtrate was quantified by measuring absorbance at 412 nm using a spectrophotometer as previously described(Omar et al., 2015, 2019, 2020).

In Vitro Time-Dependent Hb Oxygen Equilibrium Studies Using Normal Whole Blood

MMA-408, MMA-409, MMA-410, and the control 5-HMF were used to conduct time-dependent studies on Hb oxygen equilibrium using normal whole blood following previously reported procedure by our group(Omar et al., 2015, 2019, 2020). Briefly, normal blood samples (hematocrit 20%) were incubated with the compounds at 37° C for 24 h. Aliquots were taken at various time points and incubated in IL 237 tonometers (Instrumentation Laboratories, Inc. Lexington, MA) for 10 minutes at 37°C, and allowed to equilibrate at oxygen tensions 6, 20, and 40 mmHg. The samples were then aspirated into an ABL 800 Automated Blood Gas Analyzer (Radiometer) to determine the pH, partial pressure of CO₂ (pCO₂), partial pressure of oxygen (pO₂), and Hb oxygen saturation values (SO₂). The measured values of pO₂ (mmHg) and SO₂ at each pO₂ value were then subjected to a non-linear regression analysis using the program Scientist (Micromath, Salt Lake City, UT) to estimate P₅₀ as previously reported(Omar et al., 2015, 2019, 2020).

Results and discussion

Rationale and design

5-HMF is a naturally occurring non-toxic aromatic aldehyde that forms reversible Schiff base adduct with the N-terminal αVal1 amines of the two α-subunits of Hb to shift the allosteric equilibrium to the R-state Hb and increase Hb affinity for O₂, resulting in a significant sickling inhibition activity.(Abdulmalik et al., 2005; Oder et al., 2016; Safo et al., 2004; Safo & Kato, 2014) 5-HMF progressed through phase I/II clinical trials in individuals with SCD, with several positive clinical outcomes, including reduced pain, decreased lactate dehydrogenase and/or RBC hemolysis, reduction in diastolic blood pressure, and increase in SpO₂ during hypoxia challenge.(Oder et al., 2016; Pagare et al., 2022; Safo et al., 2021; Safo & Kato, 2014) However, despite promising findings, the Phase II study was terminated in part due to rapid metabolic oxidation of the aldehyde moiety(Godfrey et al., 1999; Parikh & Venitz, 2014) shortening 5-HMF’s pharmacologic effect, via a very short half-life (1–1.5 hrs) and low bioavailability. Similar metabolic instability of other antisickling aromatic aldehyde compounds (Valeresol and Vanillin) in part led to the termination of their clinical and pre-clinical development, respectively.(Pagare et al., 2022; Safo et al., 2021) Several derivatives of 5-HMF have been investigated for potential increased interactions with the protein, which was expected to increase their potency and potentially improve their pharmacokinetic (PK) properties.(Xu et al., 2017) One such potent compound is 5-PMFC, which showed about 4-fold sickling inhibition potency over 5-HMF, but without any significant improvement in its PK property compared to 5-HMF.(Xu et al., 2017)

Our group has also studied several Michael acceptor molecules that were developed to bind to βCys93 (a known allosteric site on the β-subunits of Hb),(Omar et al., 2015, 2019, 2020) based on a previously reported study with Ethacrynic acid (Kennedy et al., 1984). This interaction is expected to increase the protein affinity for oxygen and prevent hypoxia-induced RBC sickling.(Omar et al., 2015, 2016, 2019, 2020; Pagare et al., 2022). One such example is the imidazolyl acryloyl derivative, KAUS-15(Omar et al., 2015), with a β-substituted enone reactive moiety (Figure 1). This compound only weakly increased Hb oxygen affinity. Following, targeted modifications of KAUS-15 led to the development of more potent compounds, e.g. KAUS-38 (also β-substituted enone reactive moiety), a non-acidic derivative of KAUS-15 that exerted better allosteric and antisickling activities which were correlated with covalent interaction with βCys93.(Omar et al., 2019) Unlike the aromatic aldehyde 5-HMF, the Michael acceptor class of compounds provided much longer activities due to resistance to metabolic oxidation, as proven by their sustained activities in time-dependent OEC test.(Omar et al., 2015, 2019, 2020)

In this study, we took advantage of the 5-hydroxymethyl and 5-alkoxymethylfuran motifs as proven antisickling substructures and the Michael acceptor functionalities as oxidation-resistant moiety to develop novel compounds by replacing the aldehyde moiety of 5-HMF or 5-PMFC with a Michael acceptor reactive center. Eight such derivatives were synthesized (Figure 2), that we hypothesized would form covalent interaction with βCys93 on the β-subunits of Hb to increase the protein affinity for oxygen while maintaining metabolic stability of the Michael acceptor reactive center. Two classes of compounds were designed and synthesized. The first class (Figure 2, Class A) includes terminal propenoyl or propynoyl reactive moieties (MMA-409, and MMA-410) that aims to abolish the bulk effect of the imidazole ring (see side-chain in KAUS-15 and KAUS-39), which was believed to have negatively impacted the binding of the KAUS compounds within the small binding pocket surrounding βCys93.(Omar et al., 2016, 2019) The second class (Figure 2, Class B) includes non-terminal (substituted) propenone reactive moieties (MMA-401, and MMA-403), non-terminal propynone reactive moieties (MMA-407, MMA-408, and MMA-411) that are attached to less bulky methyl, amino or methoxy instead of imidazole. We rationalized those differences between the reactivities of conjugated enone and ynone systems towards cysteine (as a nucleophile) may help fine-tune the compounds antisicklingactivities.(Shankar et al., 2006) In addition to the above two Michael acceptor class of compounds, we tested their nitrile bioisostere (MMA-402, Figure 2, Class C) because the cyano group may provide a safe option for a reversibly covalent interacting moiety.(Berteotti et al., 2014) The compounds were synthesized as shown in Schemes 1 and 2, and tested for their functional and biological activities. Since our primary objective was to maintain 5HMF structural pharmacophore but with improved PK properties by replacing its aldehyde moiety with a more metabolically stable Michael acceptor reactive center, we used 5-HMF as a positive control in our studies.

Figure 2.

Newly designed conjugated Michael acceptor compounds based on 5-HMF and 5-PMFC structures

Chemical synthesis of MMA compounds

MMA-401 and 403 were prepared starting from 5-HMF by olefination of the carbaldehyde group using Horner-Wadsworth-Emmons method. This method afforded the intended trans isomers of the target compounds.(Wadsworth, 1977) The nitrile analogue MMA-402 was synthesized in two steps from 5-HMF by first condensing the aldehyde group with hydroxylamine to oxime (4) followed by dehydration with trifluoroacetic anhydride (TFAA) (Scheme 1).

The synthesis of the remaining five compounds MMA-407 to MMA-411 started by preparation of 5-PMFC from 5-HMF by applying Mistunobu reaction for aryl-alkyl ether formation.(But & Toy, 2007) The synthesis of the structurally related MMA-409 and MMA-410 from 5-PMFC was accomplished in two steps: Grignard reaction added the ethenyl or ethynyl group to the aldehyde group to give the secondary alcohols 5a-b. Oxidation of the secondary alcohol group to ketone using 2-iodoxybenzoic acid provided MMA-409 and MMA-410. The synthesis of the alkynylfuran derivatives MMA-407, MMA-408, and MMA-409 started with the installation of the ethynyl group on the position-2 of the furan ring by the reaction of 5-PMFC carbaldehyde group with 1-diazo-2-oxopropylphosphonate as described by Müller et al.,(Müller et al., 1996) to give the pivotal intermediate 7. The conversion of the alkyne 7 into the target compound started with removing the terminal hydrogen under the effect of a strong base (n-BuLi), followed by reacting the resulting carbanion with acetyl chloride or ethyl chloroformate to obtain the ketone MMA-408 or the ester MMA-411, respectively. The carbonylation of the carbanion of 7 by reacting with CO₂ afforded the carboxylic acid intermediate 8, which was easily converted into the amide MMA-407. The final compounds were fully characterized using spectral analyses (See the Experimental section).

MMA Compounds exhibit both O₂-dependent and O₂-independent antisickling activities

The primary antisickling activities of aromatic aldehydes, such as 5-HMF or Voxelotor are due to their ability to form Schiff-base interactions with the N-terminal α-chain residue αVal1 amines of liganded Hb and stabilize the R-state Hb. This leads to an increase in Hb affinity for oxygen and concentration of the non-polymer forming oxygenated HbS (O₂-dependent antisickling activity).(Abdulmalik et al., 2020; Pagare et al., 2022; Safo et al., 2021) Non-aromatic aldehydes, e.g., thiol and azolylacryloyl derivatives, similarly exhibit O₂-dependent antisickling activity but as a result of forming a covalent interaction with the Hb surface-located β-chain residue βCys93, which leads to destabilization of the T-state Hb to increase the concentration of the non-polymer forming oxygenated HbS.(Nakagawa et al., 2014, 2018; Omar et al., 2015, 2019, 2020) Recent years have seen the development of novel aromatic aldehydes, e.g., PP-14, PP-10, and VZHE-039 that increase Hb oxygen affinity to effect O₂-dependent RBC sickling inhibition and concurrently directly destabilize the HbS polymer to effect O₂-independent RBC sickling inhibition.(Abdulmalik et al., 2020; Pagare et al., 2022; Safo et al., 2021) The latter mechanism is due to the ability of these compounds to interact and perturb a surface-located αF-helix that is known to stabilize the polymer.(Abdulmalik et al., 2020; Pagare et al., 2022; Safo et al., 2021) A similar secondary O₂-independent antisickling mechanism has also been proposed for compounds that bind to βCys93 since these surface-bound molecules could prevent deoxyHbS molecules from associating.(Kennedy et al., 1984) Nonetheless, except for a limited study with Ethacrynic acid there has not been any detailed systematic study to prove this second mechanism in this class of βCys93-binding compounds.(Kennedy et al., 1984)

We expected the MMA compounds to form covalent interactions with βCys93 and to exhibit dual antisickling activities as described above. To test this hypothesis, we first tested the MMA compounds, including MMA-401, MMA-402, MMA-403, MMA-407, MMA-408, MMA-409, MMA-410, and MMA-411, as well as the positive control 5-HMF for their O₂-dependent antisickling activities at 2 mM and 5 mM, using blood suspensions from a subject with homozygous SCD (hematocrit 20%) under hypoxic conditions (2.5% O₂) at 37 °C as previously reported.(Abdulmalik et al., 2005, 2020; Pagare et al., 2018) The results are presented in Table 1, and Figure 3A. All compounds showed dose-dependent RBC sickling inhibition (9% – 64% RBC sickling inhibition at 5 mM), although lower than 5-HMF (85%). MMA409 and MMA410 were the most potent (~64%), followed by MMA-408 and MMA-401 (~44%), while the rest showed less than 25% RBC sickling inhibition. Consistent with the O₂-dependent antisickling mechanism, aliquots from the above antisickling experiment of MMA-409, when tested at 2 mM and 5 mM increased Hb affinity for oxygen by 17.7%±11.0 and 43.3% %±18.6, respectively which compares with 55.3 %±27.9 and 70.6% %±18.8 by 5-HMF (Figure 3C). The results correlate well with the trend in antisickling potency of MMA-409 and 5-HMF. Interestingly, in another oxygen equilibrium study, MMA-409 when tested with both whole blood and hemolysate at 2 mM and 5 mM showed unusually high increased in Hb oxygen affinity (11.3 %±5.6 and 49.0%±1.8, respectively) in whole blood when compared to 2.3 %±1.4 and 28.7%±3.6 with hemolysate (Figure 3D). This observation suggests that the biological effect of MMA-409, in part, includes acting on the RBC membrane. Previous studies, including a recent one by Chowdhury et al., has demonstrated that (chemically-induced) RBC shape changes affect oxygen affinity.(Chowdhury et al., 2017) We note that the activity of the MMA compounds is not unique to any of the three classes of the compounds. Among the three most potent RBC sickling inhibitors, MMA-408 is a non-terminal (hindered) ynone, while MMA-409 and MMA-410 are terminal enone and ynone, respectively.

Table 1:

Antisickling studies of MMA compounds using human sickle blood^a

Compound	% Sickling inhibition (Hypoxia)b		% Sickling inhibition (Anoxia)c
	2 mM	5 mM	2 mM	5 mM
MMA-401	16.6 ± 10.4	42.9 ± 6.5	34.9 ± 7.7	38.9 ± 8.8
MMA-402	−0.4 ± 5.1	9.1 ±4.9	20.4 ± 10.4	12.9 ± 18.9
MMA-403	9.2 ± 3.5	25.5 ±12.0	17.1 ± 14.4	23.9 ± 9.3
MMA-407	11.2 ± 4.6	24.3 ±8.2	12.0 ± 12.3	16.3 ± 17.0
MMA-408	20.8 ± 8.6	44.5 ±8.7	30.2 ± 28.5	61.4 ± 4.0
MMA-409	26.6 ±10.5	61.4 ±20.2	41.1 ± 5.2	60.2 ± 7.4
MMA-410	17.9 ± 10.7	64.1 ± 7.3	13.5 ± 24.1	54.5 ± 13.2
MMA-411	9.8 ±4.4	25.9 ± 16.7	24.6 ± 11.0	47.9 ± 6.7
5-HMFe	59.0 ±2.3	84.5 ± 3.5	−3.3 ± 4.5	−6.8 ± 12.0

^aAll studies were conducted with SS cells suspensions (20% hematocrit) incubated with 2 and 5 mM of test compounds. The results are the mean values ± SD for 3–4 individual replicate experiments. The final concentration of DMSO was <2% in all samples, including in control samples.

^bRBC sickling inhibition studies with SS cells were conducted under hypoxia (2.5% Oxygen).

^cRBC sickling inhibition studies with SS cells were conducted under anoxia condition (100% Nitrogen).

Figure 3. Biological effects of MMA compounds under various conditions.

(A) Antisickling activities of compounds under hypoxia condition (2.5% O₂). (B) Antisickling activities of compounds under anoxia condition (100% N₂ gas). (C) Effect of MMA-409 on Hb affinity for oxygen (P₅₀). (D) Effect of MMA-409 on intact red cells and hemolysate. All studies were conducted with SS cells suspensions (20% hematocrit) incubated with 2 and 5 mM of test compounds. The results are the mean values ± SD for 3–4 individual replicate experiments. The final concentration of DMSO was <2% in all samples, including in control samples.

To test our second hypothesis that the MMA compounds exhibit a second O₂-dependent antisickling activity, they were further studied under anoxia (100% nitrogen) following the hypoxia sickling studies and as previously reported(Abdulmalik et al., 2020). Remarkably, almost all compounds exhibited a dose-dependent O₂-independent antisickling activity, while 5-HMF as expected from previous studies showed no effect (Table 1 and Figure 3B). We also note that Voxelotor does not show such O₂-independent antisickling effect, although its overall potency is significantly higher than 5-HMF and the MMA compounds. This sickling inhibition under anoxia generally mirrored the effects under hypoxia, where MMA-408, MMA-409, MMA-410 showed over 54% sickling inhibition at 5 mM, followed by MMA-411 with ~48% inhibition, and MMA-401 with ~40% inhibition. The rest of the compounds showed less than 25% inhibition. Clearly, some of these compounds even though show relatively low O₂-dependent antisickling activity but make it up with their ability to stereospecifically prevent HbS molecules from coming together to form a polymer. This unique dual mechanism of action is expected to prevent sickling without drastically changing oxygen delivery capabilities and, thus, better clinical outcomes.

MMA compounds form covalent interaction with βCys93 to effect their biological activities

The atomic basis of aromatic aldehydes O₂-dependent and O₂-independent antisickling activities have been elegantly elucidated using X-ray crystallography;(Abdulmalik et al., 2011, 2020; Ahmed et al., 2020; Deshpande et al., 2018; Pagare et al., 2018, 2022; Safo et al., 2004, 2021) the former mechanism due to Schiff-base formation between the aldehyde moiety of the compounds with the N-terminal Val1 amine of liganded Hb that stabilizes the R-state Hb. The latter mechanism is due to the compounds ability to make additional interactions with the surface-located αF-helix that perturb its position, resulting in polymer destabilization. The αF-helix is important in stabilizing the pathologic βVal6 initiated polymer via intermolecular hydrogen-bond interactions that are mediated by αAsn78 or βAsp73. Consistently, SCD individuals with the classic βVal6 mutation, and a second mutation, αAsn78→Lys (Hb Stanleyville) or βAsp73→Val (Hb Mobile) on the αF-helix exhibit significantly reduced polymerization due to disruption of secondary contacts.(Benesch et al., 1979; Cretegny & Edelstein, 1993; Ferrone, 2004; Ghatge et al., 2016; Nagel et al., 1980; Rhoda et al., 1983) In the absence of X-ray crystallography (attempts to obtain co-crystals failed), we conducted two solution studies to elucidate the basis of the MMA biological/functional activities, and also confirm our hypothesis that the MMA compounds react covalently with βCys93 of Hb. The first was a disulfide exchange reaction between DTNB and the thiol of βCys93 with MMA-408, MMA-409 and MMA-410 following previously reported procedure(Omar et al., 2015, 2019, 2020). Ethacrynic acid, which is known to bind to the thiols of hemoglobin and thus limits the reaction between DTNB (5,5′-dithiobis-(2-nitrobenzoic acid) and βCys93, was studied as a positive control(Omar et al., 2015, 2019, 2020). In the presence of MMA-408, MMA-409 and MMA-410, the free thiols of βCys93 were 24%, 13% and 9% respectively (Figure 4), suggesting possible interaction of these compounds with βCys93. These results compare with the negative and positive controls, PBS/DMSO and Ethacrynic acid that showed 100% and 32% free thiols, respectively. A note of caution is that even though there are examples of compounds with Michael addition reactive moieties or sulfhydryl group targeting βCys93 to effect their antisickling activities, these compounds may exhibit non-specific binding. The apparent interaction between the MMA compounds and βCys93 is expected to abrogate the T-state stabilizing salt-bridge interaction between βAsp94 and βHis146, shift the allosteric equilibrium to the R-state(Nakagawa et al., 2014, 2018; Omar et al., 2015, 2019, 2020), consistent with MMA-409 increasing Hb affinity for oxygen. It is also likely that the O₂-independent antisickling activity of the compounds is due to binding with the surface-located βCys93 that should prevent HbS molecules coming close to each other to form the polymer. Consistently, 5-HMF, which neither binds to βCys93 nor make interaction with the αF-helix lacks this second antisickling mechanism of action.

Figure 4:

Percentage free SH (free thiol) available on Hb after reactivity of test compounds (2 mM) with βCys93 of hemoglobin (n = 2 or 3 individual replicate experiments). ECA is Ethacrynic acid.

To confirm that the MMA compounds bind to the β-subunit (and possibly to βCys93) as suggested by the disulfide exchange reaction, the representative compound MMA-409 was tested for its effect on Hb modification using aliquots from the antisickling and oxygen affinity studies as described previously(Omar et al., 2020). At 2 mM and 5 mM concentrations, MMA-409 modified Hb by 9.5%±4.6 and 26.3%±13.3, respectively, which is about 2-fold lower than modification by 5-HMF (20.8%±4.1 and 48.4%±6.4, respectively) (Figures 5A). Following, a reverse-phase HPLC study showed exclusive adduct formation with the β-chain of Hb, which contrasts with 5-HMF that binds exclusively to the α-chain of Hb (Figure 5B). Both of these studies demonstrate that the MMA compounds indeed bind to the β-chain subunit, and presumably to βCys93 to effect their RBC sickling inhibition activities.

Figure 5.

Effect of MMA-409 on Hb adduct formation in vitro (Hct of 20%) using lysates from the antisickling study. (A) Cation exchange HPLC analysis of 0 mM (control), 2 mM and/or 5 mM of MMA-409, and 5-HMF after incubation with intact SS Cells. (B) Reverse-phase HPLC analysis of 0 mM (control), 2 mM and 5 mM of MMA-409 after incubation with intact SS cells.

MMA compounds exhibit sustained biological effects compared to 5-HMF

As noted earlier, 5-HMF and other aromatic aldehydes undergo extensive oxidative metabolism by aldehyde dehydrogenase, and aldehyde oxidase in the liver, blood and other tissues, which results in poor PK properties (short half-life and suboptimal bioavailability).(Abdulmalik et al., 2005; Abraham et al., 1991; Godfrey et al., 1999; Obied &, T., Venitz, J, 2009; Safo et al., 2004; Vasiliou et al., 2000; Yoshida et al., 1998) The primary objective for developing these novel MMA compounds therefore was to improve on the PK properties of 5-HMF by replacing its aldehyde moiety with a more metabolically stable Michael acceptor reactive center, which we expected to reduce the apparent rapid oxidative metabolism, while also binding covalently to increase Hb affinity for oxygen. In an in vitro time-dependent Hb oxygen equilibrium study with MMA-408, MMA-409 and MMA-410 (at 2 mM concentrations) using freshly drawn normal whole blood and following published procedure(Omar et al., 2015, 2019, 2020), we observed that unlike the biological activity of 5-HMF which declined rapidly due to metabolism of the aldehyde, the MMA compounds showed relatively slower decline in their reactivity (Figure 6). Blood contains enzymes that are known to metabolize aromatic aldehydes and other compounds, and is therefore a good predictor of their metabolic stability.(Godfrey et al., 1999; Vasiliou et al., 2000; Yoshida et al., 1998) Although other metabolic routes, including the liver could reduce the pharmacologic activity of our novel compounds, we believe the observed improvement in the metabolic stability is significant.

Figure 6:

Time-dependent P50-shift of Hb in normal whole blood incubated with 2mM test compounds (n = 2 or 3).

Conclusion

Even though 5-HMF and several aromatic aldehydes show promise as antisickling agents, their development into therapeutic agents have been hampered by poor PK properties due in part by the metabolically unstable aldehyde moiety. We have developed novel antisickling compounds by derivatization of 5-HMF, including replacement of the aldehyde function with the metabolically stable Michael acceptor reactive center. These compounds exhibit dual antisickling activities, the first such detailed report for a different class of compounds apart from aromatic aldehydes. The compounds were designed to bind to βCys93, which is expected to lead to disruption of the T-state stabilization salt-bridge interaction between βHis146 and βAsp93, and shift the allosteric equilibrium to the R-state to increase the oxygen affinity of the protein,(Nakagawa et al., 2014, 2018; Omar et al., 2019) explaining the compounds’ O₂-depdendent antisickling effect. Additionally, because the compounds possibly bind to βCys93, which is located at the surface of the protein they are expected to weaken the interactions between adjacent HbS molecules in the polymer, consistent with their O₂-independent antisickling activities. Although not yet completely understood, this novel class of compounds may also have an additional antisickling mechanism by acting on the RBC membrane as evidenced by oxygen equilibrium studies with MMA-409. Generally, this class of compounds showed improved in vitro metabolic profile when analyzed for time-dependent effects on Hb oxygen affinity using fresh whole blood. We anticipate that these unique properties of the MMA compounds can be further optimized structurally, with guidance from the putative βCys93 binding pocket, and would translate into sustained and improved pharmacologic activities in vivo without drastically changing oxygen delivery capabilities.